Gary Taubes on nutrition epidemiology as a basis for guidelines: warns against

Gary Taubes warns that observational nutrition studies — food-frequency questionnaires and the associations drawn from them — cannot establish cause and effect and should not drive national dietary recommendations. He made the case in a January 1, 2026 Levels podcast appearance.

His stated reasoning: he argues that epidemiological studies in nutrition are riddled with measurement error and confounding, that their weak associations are routinely presented as causation, and that building guidelines on them means building policy on unreliable foundations. He treats this methodological critique as the root of much of the field’s error: if the evidence base cannot support causal claims, the advice built on it should be held to a higher standard of humility.

This page chronicles a stated position of the person above. It is not medical advice and not the site's own conclusion.